Engaging Overseas Sri Lankans to Facilitate Export Diversification

Jack Sennett

This is an edited version of a Policy Analysis written in fulfilment of the requirements for the degree of Master in Public Administration in International Development, John F. Kennedy School of Government, Harvard University

March 2018

Executive Summary

Insufficient export diversification is a binding constraint to economic growth in Sri Lanka

• The Harvard CID growth diagnostic found that with wages in traditional export sectors now below average Sri Lankan wages, new higher-wage export industries are required

Overseas Sri Lankans (OSL) have the potential to create new export industries in Sri Lanka

• Diasporas were involved in the export-led development of India, Taiwan, and China by bringing industry knowhow and market connections to their home countries • There are large, well-educated OSL communities living in the US, UK, Canada, and Australia that have the industry knowhow to assist in export-led growth in Sri Lanka

OSL can have the biggest impact on diversifying exports if they return to start firms in new export industries rather than working with firms while based overseas

• OSL can play a useful role connecting the existent Sri Lankan IT export sector to overseas markets, but they cannot start firms in new export industries from abroad • If OSL return to start firms they can “seed” a new export industry that grows organically through the diffusion of knowhow • The pharmaceutical sector is an example of an industry with high potential to be “seeded” by returning OSL entrepreneurs

Preliminary policy recommendations focus on removing barriers and catalyzing latent motivations to facilitate OSL return entrepreneurship:

• The Department for Immigration and Emigration should continue to ease border processes for OSL through dual citizenship and the OSL lifetime resident visa • The Board of Investment should orient part of its “one-stop-shop” to dealing specifically with OSL issues • The Ministry of Foreign Affairs should utilize its diplomatic network to engage potential OSL entrepreneurs to catalyze latent motivations to return

Table of Contents

Executive Summary ......

1. Problem Outline ...... 1

2. Engaging with diasporas: a theoretical review ...... 2

a. The definition and origins of the Sri Lankan diaspora and the introduction of the term Overseas Sri Lankans (OSL) ...... 2

b. Diaspora as a driver of export growth ...... 3

i. The “return option”: knowledge transfer and industry “seeding” ...... 3

ii. The “overseas option”: investment, knowhow diffusion, and trade links...... 6

c. Analytical framework ...... 9

3. Characteristics of Overseas Sri Lankans ...... 13

a. Where do Overseas Sri Lankans live? ...... 13

b. How does the OSL community differ between countries? ...... 15

c. OSL Country Profile: United States ...... 16

d. OSL Country Profile: Australia ...... 18

e. Conclusion ...... 20

4. Case Studies of OSL engagement with Sri Lanka ...... 21

a. Return option ...... 21

b. Overseas option ...... 23

5. Lessons for Policy from Current OSL Engagement ...... 25

a. Motivations and Catalysts ...... 26

b. Barriers ...... 28

6. Return Option vs Overseas Option ...... 31

7. Preliminary policy recommendations ...... 36

a. Ease OSL dual citizenship acquisition...... 36

b. Visa reform ...... 37

c. OSL one-stop-shop ...... 38

d. OSL engagement through embassies, high-commissions, and consulates ...... 38

Appendix A – Interview list ...... 41

Appendix B – Pharmaceutical outsourcing ...... 42

Appendix C – OSL Characteristics in USA and Australia ...... 43

Appendix D – Entrepreneurially likely occupation categories in Australian census data ...... 46

Appendix E – Additional case study ...... 46

Appendix F – De-prioritized preliminary policy recommendations ...... 48

a. Capital controls ...... 48

b. OSL celebration events ...... 48

References ...... 50

iii

1. Problem Outline

a. Export diversification as a constraint to growth The Center for International Development at Harvard University in collaboration with the Millennium Challenge Corporation conducted a growth diagnostic for the Prime Minister’s office of Sri Lanka (Center for International Development, 2018). The growth diagnostic concluded that the core constraint to growth was the widening current account deficit. CID argued that to reduce the deficit, exports need to grow to keep pace with rising imports, and that to enable this, the Sri Lankan economy must diversify from its current traditional set of products (tea, rubber, and garments). These traditional industries are no longer able to pay wages commensurate with Sri Lanka’s level of development, and new higher-wage export industries are required. How the government enables this diversification remains a big question, and this study intends to help answer that question. b. Diaspora as a source of export relevant knowhow The growth diagnostic identifies several binding constraints to diversifying exports related to infrastructure and government failures: access to land, policy uncertainty (especially tax and trade) and wastewater and transport infrastructure. Alongside these constraints however there is a market failure in self-discovery. While firms in Sri Lanka have done well to enter new product lines in industries similar to the current export base (e.g., different kinds of garments) they have not been able to “jump” to new export industries. Too few firms are investing in new industries, thereby enabling Sri Lanka to discover which export industries it has a comparative advantage in. Tackling the infrastructure and government failures mentioned will incentivize FDI and local firms to invest, but it is also possible to tackle the market failure directly by increasing the inflow of knowhow relevant to new export industries.

Sri Lanka has a large overseas population, many of whom live and work in advanced economies that are rich in knowhow. The Overseas Sri Lankans (OSL) in these (mostly English-speaking) advanced economies represent a large source of tacit knowhow and connections to foreign markets that could flow to Sri Lanka through return migration or personal ties. The study aims to understand the current knowhow that resides in OSL communities and how to assist its flow

to the Sri Lankan economy, thereby helping Sri Lanka discover its next competitive export industry.

Engaging Overseas Sri Lankans brings many economic and non-economic benefits other than export diversification, such as cultural exchange, remittances, and family support. This scope of this paper however is limited to the benefits of engaging OSL related to export diversification.

2. Engaging with diasporas: a theoretical review

a. The definition and origins of the Sri Lankan diaspora and the introduction of the term Overseas Sri Lankans (OSL) It is useful to first to ask the question of whether a Sri Lankan diaspora exists, and what character it takes. The work of Robin Cohen on Global Diasporas can aid us here (Cohen, 2008). Certainly, there are many people who were born in Sri Lanka (or whose ancestors were) and who live outside of Sri Lanka, however Cohen asserts that more is required to define a diaspora. Table 1.1 in (Cohen, 2008) outlines common features of diaspora, of which the Sri Lankan diaspora confirms with at least three (italics mine):

1. Dispersal from an original homeland, often traumatically, to two or more foreign regions (Sri Lankans have moved to many foreign regions in North America, Europe, the Middle East and Australia, which for many Tamil Sri Lankans was due the civil war) 2. A collective memory and myth about the homeland, including its location, history, suffering and achievements (Sri Lanka has a strong national identity with well documented historical achievements e.g., ancient cities of Sigiriya and Kandy) 3. A strong ethnic group consciousness sustained over a long time and based on a sense of distinctiveness, a common history, the transmission of a common cultural and religious heritage and the belief in a common fate (OSL continue to gather together in diaspora organizations, suggesting there remains a strong ethnic group consciousness, although this is sometimes divided between Tamil, Sinhalese, and Muslim communities)

Few would dispute the existence of Sri Lankan diaspora communities that conforming to the above features, which leads us to the questions of what kind of diaspora is the Sri Lankan

diaspora. Cohen defines five typologies of diaspora: Victim (dispersal following a traumatic event in the homeland), labor (emigration in search of work), Imperial (emigration for further colonial ambitions), Trade (emigration in search of business and trade) and Deterritorialized. Different Sri Lankan diaspora communities fall into different categories. Many Tamil communities could be described as Victim diaspora, forced to leave their homes and country during the 30-year civil war. Tamil, Singhalese, and Muslim diaspora fall into the labor category, especially in the Middle East, as well as the trade category, with a large contingent of highly educated Sri Lankans studying and working in professional and business occupations in high- income countries.

Assigning each member of the Sri Lankan diaspora into one of five typologies would be an oversimplification (e.g., some Tamils may fall into both trade and victim typologies), nonetheless some demarcation of diaspora communities provides a framework to focus on the communities most likely to engage economically with Sri Lanka. Given the purpose of export diversification, the most promising is the trade diaspora.

Finally, the use of the term diaspora is interrogated. Consultations carried out by the Young Political Forum found that some Sri Lankans believed the use of the term ‘diaspora’ to describe Sri Lankans living overseas had negative connotations and was largely associated with the Liberation Tigers for Tamil Eelam (Young Political Leaders’ Forum and International Alert, 2016). The Ministry of Foreign Affairs appears to have agreed with this view and has changed the term to Overseas Sri Lankans (OSL). This report will follow that recommendation and use the term OSL instead of diaspora.

b. Diaspora as a driver of export growth

i. The “return option”: knowledge transfer and industry “seeding” A well studied channel through which diaspora can contribute to economic diversification in their country of origin is return migration – the “return option”. (Borjas, 2014) outlines a theoretical model of return migration where the introduction of non-rivalrous ideas from immigrants increases the marginal productivity of all other workers and enables endogenous growth. Empirical work by (Hausmann and Nedelkoska, 2017) from Albania show positive

results of return migration of diaspora from Greece in the aftermath of the financial crisis. Return diaspora engaged in self-employment and entrepreneurship which led to increases in wages for non-skilled migrants and positive effects on overall employment.

Both model and empirical work focus on effects on already existent industries. If Sri Lanka is to establish new industries however, such models may not be appropriate. (Rhee and Bélot, 1990) surveyed 11 successful exporting industries in low-income countries and found that the key aspect to their development was the presence of a catalyst with the right knowhow to initiate the industry. (Mostafa and Klepper, 2017) extended this idea, theorizing that the key ingredient that entrepreneurs in low-income countries require is tacit “organizational knowledge” from more advanced economies. While embodied technical knowhow can be learned from manuals and blueprints, organizational knowhow cannot be codified, and thus is tacit. Such knowhow can only be quickly attained through on-the-job training, working alongside someone who has already mastered it.

Mostafa and Klepper explore this idea by studying the growth of the Bangladesh garments industry, tracing its origins from the Bangladeshi manufacturer Desh. They find that a technical partnership between Daewoo of South Korea and Desh was the catalyst for an entire garments industry. Daewoo trained 126 of Desh’s workers over 6 months in South Korea, providing the necessary tacit knowhow. They conclude that the ‘seeding’ of this organizational knowledge in the Bangladesh garments sector was critical to the industry’s expansion. Another example was the founding of Taiwan’s Integrated Circuit (IC) design and manufacturing industry. In 1976 the Industrial Technology Research Institute (ITRI), a government agency, sent 37 engineers to the US for a year to work with RCA, an electronics company, to enable the transfer of their semiconductor technology to Taiwan. Becoming known as the RCA-37, the group of mostly Taiwanese engineers were critical the founding the Taiwanese IC industry (Saxenian, 2006).

Historical studies of the growth of the IT industry in India, China, and Taiwan by AnnaLee Saxenian demonstrates how diaspora can play this “seeding” role using tacit knowhow they have acquired in the high-income countries where they live (Saxenian, 2006). While much of the early industrialization in Taiwan was state and MNC led (such as the IC industry) she

describes how returning diaspora became more important. For example, the scanner industry in Taiwan was founded by returning Chinese entrepreneurs who had worked together at Xerox and came back to Taiwan in 1980 to start a company called Microtek. They built a supply chain of scanner components around their product by nurturing local subcontractors and suppliers. Soon, local competitors opened up and by 2000 Taiwan produced 90% of the world’s scanners, with Microtek itself producing 20% of the market. Microtek’s founders were early movers, but partly inspired by the success of such firms and facing a US technology recession at the beginning of the 1990s, 40,000 Chinese returned to Taiwan from the US between 1986-1996. By the year 2000, 42% of company founders located in the Hsinchu science park (the home of the Taiwanese high-tech sector) were diaspora returnees (Saxenian, 2006).

The nascent beginnings of an export IT industry in India in the 1980s was largely driven by multi-national corporations (MNCs) opening offshore centers in India, but it was often returning Indian diaspora that opened the offices. HP-India, created in 1985, was the first of two offshore centers in India and was set-up by Radha Basu, an Indian diaspora who had been educated in the US and worked as a manager at HP in America (Saxenian, 2006). Alongside the MNC expansion there were also diaspora doing it themselves. For example, an IT hardware industry was started when Azim Premji returned home in 1979 to take over the family cooking oil business. Having completed an engineering PhD at Stanford, he re-orientated the business towards IT hardware, renaming in Wipro. Eventually Wipro moved into the software business and after recruiting another US-based Indian diaspora in 1999 from General Electric to spearhead the new software services segment, grew into the $8bn company it is today.

Diaspora played another major role in India after the dot-com crash in 2001: “some 5,000 overseas Indian technology professionals with more than five years of work experience moved back to India from the US in 2002-3” (Saxenian, 2006). Again, there were many who went back to set up operations for MNCs like the founding of Intel’s new microprocessor manufacturing plant in 2002. Some however started their own firms such as the integrated circuit manufacturer Ittiam Systems that was founded in 2001 by Indian diaspora who had been senior managers at Texas Instruments.

Industry seeding has strong theoretical and empirical foundations, including through the return of diaspora from more advanced economies. It presents a promising avenue for OSL to enable export diversification in Sri Lanka.

ii. The “overseas option”: investment, knowhow diffusion, and trade links So far, I have considered what (Samet, 2013) coins as the “return option”. He also however outlines a “diaspora option”, which, in light of the previous discussion of the term “diaspora”, I call the “overseas option”. Samet defines this option as: “creat[ing] channels through which the expatriates would be effectively and productively related to the development of their country of origin, without any temporary or permanent physical return” (Samet, 2013). We can think of the overseas option as spending no more than 3 months a year in Sri Lanka (but typically much less than that) and the return option as spending at least 6 months a year in Sri Lanka for a significant period of time (i.e. at least 1-2 years). Samet describes different channels of this economic engagement including (a) mobilization of capital for investment, (b) the creation of diaspora networks that facilitate knowhow diffusions back to the country of origin, and (c) greater links for trade and Foreign Direct Investment (FDI).

(a) mobilization of capital: Diaspora are seen as more likely to mobilize and invest capital than other foreign citizens because they have a lower perception of risk of investing in their country of origin. This can be because of familiarity with the culture reduces information asymmetries, and because exchange rate fluctuations are less problematic if they have financial obligations in their home country (Agunias et al., 2012). This might vary by diaspora typology, with victim diaspora perceiving higher risk of investing in their country of origin than e.g., trade diaspora.

(b) knowhow diffusion: Arguments for knowhow diffusion are similar to those for the “return option” with the cultural links enabling knowhow to flow to the country of origin.

(c) trade links: A theoretical grounding for diaspora to facilitate trade links was outlined by (Greif, 1993). He uses game theoretic models to explain how 11th century Maghribi traders used informal networks to enforce contractual norms and transmit information. While the US may have formal enforceable contract law that negates such a network’s advantage, Sri Lanka has a much less formalized business environment. (Hallward-Driemeier et al., 2010) describe how

6 low-income countries often face deals rather than policy rules. Sri Lanka is ranked 165/190 for enforcing formal contracts in the World Bank’s doing business indicators (World Bank Group, 2017), indicating it probably has a deals-based business environment. This suggests there is an economic advantage to informal contractual enforcement in cross-border economic activities that involve the US and Sri Lanka. OSL can play this role as they both understand the formal rules of the US, while also maintain family and informal networks in Sri Lanka that enable them to create and enforce deals that US nationals would not be able to (see box on game theoretic analysis for further details).

Box: Game theoretic analysis of OSL facilitating exports in the overseas option

Following (Greif, 1993) I can model an advantage of diaspora connections where Sri Lankan entrepreneurs lack sales knowhow and contacts within export markets such as the US. Forming partnerships with US-based sales agents may not be possible due to the difficulty of enforcing contracts across national borders (US agents cannot enforce contracts in Sri Lanka due to the deals environment and Sri Lankans face prohibitive costs and administrative difficulties enforcing contracts in the US). US-based sales agents cannot be trusted by Sri Lankan entrepreneurs not to cheat and vice versa. This may be exacerbated by the difficulty of monitoring across borders, especially with social, cultural and language barriers.

Business is a repeated game, and thus the cost of cheating must be increased enough such that the rewards of cooperating over time make cooperating the dominant strategy for both players. Social ties between OSL and Sri Lankans may solve this problem by providing a sufficiently high cost should either sales agent or entrepreneur cheat the other. Should Sri Lankan entrepreneurs cheat their OSL sales agents, they may gain a reputation as a cheater in Sri Lanka that prevents them engaging in profitable commercial transactions in the future. Given the relatively small business community in Sri Lanka this is not infeasible. This could represent a very large forgone benefit for entrepreneurs that along with the benefits of continued business with the sales-agent far outweighs the short-term benefits of cheating.

OSL face a slightly different trade-off however, since their US-based commercial opportunities are not likely to be affected by a loss of reputation amongst Sri Lankan communities. Enforcing

contracts on US-based sales agents is likely easier however, and OSL are probably more accustomed to a rules economy, reducing the likelihood of cheating. The loss of reputation also has non-monetary social costs amongst their family in Sri Lanka and the wider OSL community. The social links between OSL and Sri Lankans also enable a greater monitoring capability for both sides of the business. This reduces the rewards of cheating since either side will be able to more easily detect cheaters earlier. Thus, economic transactions which may not have had a sustainable Nash equilibrium with non-OSL US-based sales agents find a Nash equilibrium when the payoffs are modified by the sales agent’s status as an Overseas Sri Lankan.

There is quantitative empirical evidence to support the theoretical model of diaspora contributing to an increase in trade between their country of origin and country of residence. A study in Canada found that increasing a diaspora population by 10% resulted in an increase in imports from their country of origin by 3% (Head and Ries, 1998). This is particularly encouraging since one of the largest OSL communities is in Canada and other large diasporas are in countries of a similar level of economic development (e.g., US, UK, Australia).

There is also historical evidence that shows that diaspora taking the overseas option can be important to a countries’ export growth. As noted earlier, returning diaspora were important for the India IT export industry in the 80s and early 2000s, but the overseas option played a major role in 1990s. This key phase in the growth of India’s IT sector was facilitated by Indian diaspora in the US “mentoring early-stage companies and confidence building with major U.S. corporations (that India was a good place to get work done and Indian companies have the required wherewithal to perform the work)” (Pandey et al., 2004). Similarly, in Taiwan, global IT firm Acer was started in 1976 by local Taiwanese but was assisted by connections to the Chinese community in Silicon Valley. They opened at office in the US in 1977 to help source components and keep abreast of advances in technology (at this time they were serving the domestic Taiwanese market) and recruited a former classmate of theirs who then lived in Silicon Valley to run it. The Co-founder George Huang recalls that “it was impossible to lure Chinese engineers back to Taiwan during the 1970s, but they were an essential source of knowledge and contact for their friends and classmates in Taiwan” (Saxenian, 2006).

c. Analytical framework

i. Channels of economic engagement Figure 2 takes the theoretical channels of economic engagement discussed above and integrates them into a framework of options facing an OSL looking to engage with Sri Lanka. It uses the return and overseas options as the initial start point. The channels most relevant for export diversification are highlighted.

Figure 2: The two most relevant channels of engagement are return and overseas entrepreneurship

I choose to exclude several channels with a low likelihood of affecting export diversification. The employment channel is excluded since Sri Lanka needs to create new companies in absent or nascent industries if it is to diversify its exports. This cannot be achieved purely by gaining employment in an already existing firm (I exclude from this setting up an office for a MNC).

Investment is excluded on the basis that access to capital was not deemed a binding constraint to growth in the CID growth diagnostic (Center for International Development, 2018). There is a lack of investable opportunities in the export sector rather than a lack of capital to invest in such opportunities. Paid advisor roles are also excluded for similar reasons to employment, that the shortage is sufficient new companies in new sectors, not individuals willing to work or provide advice for existing companies. Advisors that require little compensation could assist in starting a new company (when there is unlikely to be funding for salaries) however an advisor who occasionally visits Sri Lanka from overseas is very unlikely to be able to transfer the multitude of tacit knowhow required to organize a new company in a new industry. Lastly, civil society and NGO entrepreneurship is excluded, as while no doubt valuable for many reasons, they do not directly tackle this issue of export diversification.

The highlighted options both fall under the category of “Entrepreneurship”. The definition of entrepreneurship should be taken quite broadly, encompassing actions that are not necessarily founding a company but at least starting some novel aspect of it (using capabilities not present in the rest of the company) e.g., running a new operations center in Sri Lanka for an MNC. For the overseas option, the private sector entrepreneurship channel is limited to two categories: offshoring and sales agents. Offshoring refers to the act of an OSL who works for an overseas firm, initiating the creation of a Sri Lanka office or manufacturing plant that offshores activities for the firm (now an MNC) but not returning permanently to run it. The sales agent option refers to the classic concept of a sales agent e.g., an OSL that opens an office to market products on behalf of a Sri Lankan firm (and usually takes some equity). It also encompasses other less obvious sales roles such as a specialist outsourcing firm which has most of its operations in Sri Lanka, but is founded by an OSL who resides overseas and who acts as a “sales agent” of the outsourcing product. No further breakdown is given for return option private sector entrepreneurship since it is more flexible and appropriate for most business activities. ii. Highest potential industries and sectors for export diversification Both channels above require that the OSL has some expertise in the sector (either to sell the product, or to start a company that produces it). Many industries that OSL have expertise in may not be appropriate or competitive in the Sri Lankan context however. It is unlikely for 10

example that complex aerospace manufacturing, of the kind conducted by Boeing in the US and Airbus in Europe, would be possible in Sri Lanka regardless of whether there were OSL entrepreneurs with the knowhow to run such a company.

The Board of Investment and the Export Development Board in Sri Lanka in collaboration with Harvard’s Center for International Development ran a sector prioritization to determine which export sectors were most promising for Sri Lanka (Malalgoda et al., 2018). The results are summarized in table 1. This enables the government to focus on engaging OSL with the knowhow most promising for Sri Lanka. Another prioritization was done for more detailed subsectors (e.g., screw manufacturing). These industry lists are used to identify the number of OSL with knowhow relevant for the Sri Lankan economy and for informing policy recommendations where the types of industry being promoted affects how one engages with OSL.

Table 1: The BOI identified 16 high potential exports sectors for the Sri Lankan economy

Tier 1 Priority Sectors Tier 2 Priority Sectors Highest market opportunities, investor High market opportunities, investor interest (Direct & Indirect) and impact interest (Direct & Indirect) and impact

Existing and • Accommodation and food service • Software and IT Emerging (tourism) • Transportation and storage (logistics) Sectors • Food products • Finance and insurance Already showing • Rubber and plastics products strength or • Education related knowhow in SL Possible New • Electronics • Cement, ceramics, glass, and other Sectors • Industrial machinery and equipment mineral products Average to Very • Transport equipment (motor vehicles, • Pharmaceutical products Low current trailers) • Transport equipment (ships, strength & • Electrical equipment motorcycles/bicycles, other) related • Fabricated metal products knowhow, but • Other products (medical, decorative, required assets recreational) are competitive • Business administration & support (BPO) Source: (Board of Investment of Sri Lanka et al., 2017)

iii. Review of diaspora engagement policy research The IOM published a comprehensive guide to diaspora engagement (Agunias et al., 2012). They argue that in order to facilitate the return option the right mix of incentives must be in place. An obvious incentive is financial payments, however the IOM point out that “that paying diaspora members three to four times as much as locals is unfair” and may cause resentment among local population. Given the large wage differences between the western countries and Sri Lanka it is unlikely such incentives would be appropriate.

The IOM do give suggestions for reducing barriers for return migration, principally that of creating flexible citizenship and residency visa options. They also suggest granting political voting rights to overseas citizens, providing special property rights for diaspora, and introducing laws that formally recognize the diaspora, could all reduce barriers to diaspora return.

A recent report called The Roadmap for Engaging Sri Lankans Overseas (“The Roadmap”) suggests encouraging OSL to engage with Sri Lanka by holding an annual festival to celebrate the achievements of OSL (Young Political Leaders’ Forum and International Alert, 2016). This is something the Ministry of Foreign Affairs has considered and has potential applicability to both the overseas and return options. iv. Current GoSL OSL outreach policies The GoSL has recognized the potential of the OSL community to benefit the country’s development and has engaged in several policy directions to encourage this:

• Changing the terminology from “diaspora” to Overseas Sri Lankans (rationale described above) • Introducing dual citizenship: Many OSL lost their Sri Lankan citizenship when they took citizenship of another country. The government now allows dual citizenship. Any former Sri Lankan resident can apply for dual citizenship (assuming their current country allows it) if they fulfill the academic qualifications (min 1-year diploma), are over 55, or can bring approximately $25,000 into the country. Anecdotal evidence suggests former citizens are managing to claim dual citizenship (interview P), but the process is somewhat lengthy and bureaucratic (Young Political Leaders’ Forum and International Alert, 2016).

• Introducing OSL visa cateogory: In March 2018 it was announced that the government would be offering lifetime permanent residence visas to OSL who have become citizens of other countries (Wickremasekara, 2018), enabling OSL to work in Sri Lanka without applying for dual citizenship. The visa will incur a fee of $1,500, which should not put off most OSL, but the final details are yet to be released (e.g., land ownership rights). • Supporting OSL focused conferences in Sri Lanka: The NGO International Alert hosted conferences in February 2015 and October 2016 on engaging Overseas Sri Lankans, the latter of which was “informally supported” by the Ministry of Foreign Affairs (MoFA) (International Alert, 2016). The MoFA also collaborated with the Centre for Poverty Analysis (CEPA) to invite OSL to a roundtable on “Overseas Sri Lankans as Partners of Development” in April 2017 (“Roundtable on ‘Overseas Sri Lankans as Partners of Development,’” 2017).

3. Characteristics of Overseas Sri Lankans

a. Where do Overseas Sri Lankans live? Figure 3: OSL in “Western” countries are better placed to facilitate export diversification

Emigration is a part of life in Sri Lanka. The UN estimates that there were 1.64 million Sri Lankans living overseas in 2015 (United Nations, Department of Economic and Social Affairs, Population Division, 2015). It is useful to distinguish two different archetypes of diaspora that account for 90% of the OSL (figure 3). OSL in “western” countries tend to be employed in higher skill professions, whereas those in the Middle East are often hired for unskilled professions such as housekeepers (Sri Lankan Bureau of Foreign Employment, 2016).

The analytical framework outlined in the previous section outlined how OSL could create export diversification through returning with knowhow of new industries, or by creating access to large export markets. OSL in “western” countries are better suited for both these channels of engagement since the markets they live in are larger and richer in knowhow, and the OSL tend to be higher skilled. This report thus focuses on “western” OSL communities.

Figure 4: OSL are concentrated in four English-speaking countries

"Western" Overseas Sri Lankan Population by country of residence (000s)

Canada 147 UK 139 Australia 112 Italy 85 USA 48 France 48 Germany 47 Switzerland 42 Other 51

Source: (United Nations, Department of Economic and Social Affairs, Population Division, 2015)

Figure 4 shows the “western” OSL are primarily focused in English-speaking countries (446,000 live in Canada, UK, Australia, and USA) with the remaining population based mostly in four other western European countries (Italy, France, Germany, and Switzerland).

b. How does the OSL community differ between countries? As noted earlier, Sri Lankans emigrated for a variety of reasons, and these can be categorized using the typology developed by (Cohen, 2008) and described above. It is worth understanding if certain typologies are concentrated in particular countries to better provide targeted policy responses.

Identifying precisely the OSL that fled conflict (“victim diaspora”) is not possible, however since the majority were of Tamil ethnicity we can see which countries have large Tamil populations as a guide. Estimates of the size of the Tamil diaspora vary but Human Rights Watch estimates that 200,000-250,000 are in Canada, 150,000 in the UK, 50,000 in Germany, and 30,000 in France, Switzerland, and Australia (Becker, 2006). While these estimates do not match entirely with the UN estimates (e.g., they only estimate there are 148,000 Sri Lankans in Canada, possibly due to the exclusion of second generation OSL) they do indicate which countries have high populations of OSL that led Sri Lanka due to conflict. Canada, UK, France, Germany and Switzerland OSL communities contain large proportions of such OSL.

OSL in Italy are of mostly Sinhalese ethnicity, are largely employed in the domestic sector (Henayaka-Lochbihler and Lambusta, 2004), and can be categorized as “labor diaspora”. This appears to have been facilitated by a large migration in the late 70s and early 80s when Italy had more favorable immigration policies than other western European countries, and thus attracted many Sri Lankans who had failed to gain entry to the UK, Germany, Switzerland or France (Henayaka-Lochbihler and Lambusta, 2004). This is consistent with a large low-skill migration profile of Sri Lankans in Italy, who having failed to gain admission in stricter jurisdictions, travelled to Italy, and once established, increased their numbers through family reunification.

Identifying the “trade diaspora” is not straightforward but education levels can provide a good proxy. Figure 5 shows the proportion of high educated OSL in the countries with the largest OSL populations. The large proportions of highly educated OSL are based in Canada, UK, Australia and the US. The proportion of highly educated is highest in Australia and USA (57% and 56% respectively). This is reflective of the Australian points-based immigration policy, which makes

immigration significantly easier for highly educated and high-skilled workers (Interview J). It is also likely a consequence of Australia and USA being less common destinations for OSL who were fleeing conflict, and thus contains a higher proportion of OSL who left for high-skill employment and entrepreneurship opportunities. These four English-speaking countries are the most likely places to find the trade diaspora that have the knowhow and connections for export diversification. This report focuses on Australia and United States as these had highly convenient data sets, however further work profiling Canada and the UK would be highly beneficial. Below I present country profiles of the US and Australia that details the location, occupations, and industries of OSL. Understanding where OSL live, and what industrial knowhow and connections they have, is vital to formulating policy that efficiently engages the OSL with the highest potential to facilitate export diversification.

Figure 5: OSL in the four English speaking countries are very well educated

Education Levels of OSL by country of residence

Canada 41%

United Kingdom 47%

Australia 57%

Italy 4%

United States 56%

Germany 6%

France 12%

Switzerland 10%

0 20 40 60 80 100 120 140 Thousands Highly Educated Low and Medium Educated

Source: (OECD, 2015) – percentages indicate the proportion of OSL that are highly educated for each country c. OSL Country Profile: United States The American Community Survey (ACS) 2016 provides data on OSL in the US (Ruggles et al., 2017). The survey identifies approximately 39,000 first generation OSL living in the US (this is consistent with the OECD 2015 study and not dissimilar to the UN database).

OSL are mainly located in the wealthiest, most urbanized parts of the US. 27% of the population is in California (1/3 in LA), with another 20% in the New York City tri-state area. Remaining large populations are in Texas (especially Houston), Washington DC-Baltimore area, and Massachusetts. A full breakdown is provided in table 6, Appendix C.

i. What industries and professions are OSL engaged in? We saw that the OSL population in the US is a highly educated group and perhaps unsurprisingly the average adult income for OSL is $56,256 which is approximately 40% higher that the US average of $39,941 (ACS 2016). This is reflected in the typical professions of OSL in the US with high concentrations in Management, Higher Education, Computer and Mathematical, and Healthcare occupations and fewer in Personal Care or Food Preparation. The high proportion in Management and Computer and Mathematical occupations as well as significant numbers in Finance, Science and Business Operations is encouraging as these are occupations more likely to develop the knowhow for entrepreneurship in their particular industry. A full breakdown is provided in figure 6, Appendix C.

Unfortunately given the limited sample size of the ACS it is not possible to see which industries have the highest concentration of OSL in Management occupations (a group likely to engage in entrepreneurship). Table 2 presents OSL employment by industry without distinguishing by occupation. Encouragingly, 49% of OSL are employed in industries which (at the highest level) were identified as high potential by the Board of Investment (BOI). It should be noted that these industries are given at the least level of detail. For example, manufacturing has a very diverse set of subsectors within it, some of which may not be suitable for the Sri Lankan economy (e.g., Aerospace) and thus we should not conclude that half of OSL could potentially start a business in Sri Lanka. It is encouraging however that the OSL are concentrated (at a high level) in industries that the BOI identified as high potential for Sri Lanka.

Table 2: There is a high concentration of US-based OSL working in priority export industries for Sri Lanka (in bold) Industry Population Percentage Priority of total industry Educational Services 6,266 20% Yes Health Care and Social Assistance 5,553 18% No Retail Trade 3,723 12% No Professional, Scientific, and Technical Services 3,372 11% Yes Finance and Insurance 2,343 7% Yes Manufacturing 1,833 6% Yes Accommodation and Food Services 1,555 5% Yes Other (11 industries) 6,771 22% No Adult populations of OSL by industry of employment in the US, Source: American Community Survey (ACS) 2016 Bold face indicates priority industry as identified by the (Board of Investment of Sri Lanka et al., 2017). d. OSL Country Profile: Australia The Australian Census of 2016 provides detailed information on the OSL community in Australia. It identifies 109,850 Australian residents that were born in Sri Lanka, matching very closely to the UN’s estimate of the OSL population.

Australia OSL are heavily concentrated in Melbourne and Sydney with approximately 50% in Melbourne and 25% in Sydney. The remaining populations are mostly in the cities of Perth, Brisbane, Adelaide or Canberra (Australian Capital Territory). Very few live in rural areas. Table 7, Appendix C shows the breakdown for the largest centers of OSL populations.

i. What industries and professions are OSL engaged in? Like in the USA the average income of OSL in Australia is higher than that of Australia as a whole. The gap does not appear as large as the US (unfortunately there are no mean income statistics for a direct comparison) but 48% of OSL have a personal income in the top half of the income brackets defined in the census (over AUD41,600 per annum) compared with only 42%

1 of the Australian population as a whole0F . As in the US this is again reflected in a high proportion of OSL in professional and managerial occupations, with 43% of the employed OSL population

1 Excluding those that did not declare income or was not applicable. 18

working as managers (7,000) or professionals (21,000). A full breakdown is given in figure 7, Appendix C.

Table 3 shows the industries that OSL are employed in. As in the US there is a significant proportion of OSL in priority industries with 39% of those in employment working in such industries. Compared to the US there is a higher proportion in health care, and a smaller proportion in education. In general, the OSL community is employed more evenly across the different industries than in the US.

Table 3: There is a high concentration of Australia based OSL working in priority export industries for Sri Lanka (in bold) Industry Population Percentage Priority of total industry Health Care and Social Assistance 12,083 19% No Professional, Scientific and Technical 6,495 Yes Services 10% Manufacturing 5,833 9% Yes Retail Trade 4,676 7% No Public Administration and Safety 4,642 7% No Education and Training 4,467 7% Yes Financial and Insurance Services 4,088 6% Yes Accommodation and Food Services 3,677 6% Yes Other (11 industries) 17,236 27% No Source: Australian Census 2016

An advantage of the Australian Census is that it is possible to get much more granular statistics since it captures the whole population, not just a sample. I can therefore determine how many OSL are in entrepreneurially likely roles within priority BOI sectors.

The most likely profession categories which could lead to entrepreneurship are professionals and managers. Using more granular occupation definitions we can further define professions most likely to engage in entrepreneurship as managers and certain professionals (for example excluding Legal Professionals but including Business professionals). The full list of excluded and included occupations can be seen in Appendix D. Similarly, we can take more granular industry definitions to more closely match to the priority sectors identified by the BOI.

Table 4 shows that 11,000 OSL are engaged in entrepreneurially likely occupations in priority industries (shown in more granular detail than Table 3). This an encouragingly high proportion of the 63,000 OSL in employment in Australia. There are significant numbers of entrepreneurially likely OSL in Professional, Scientific and Technical, Computer System Design, and Finance and Insurance. Education is another sector with a large OSL manager and professional population, and there is ~800 or so food and beverage and accommodation OSL that could be involved in the tourism industry in Sri Lanka. There is also a welcome number of potential entrepreneurs in manufacturing, with over 1000 OSL in priority manufacturing sectors. 50% of these are in food product or machinery and transport equipment manufacturing (further breakdown is available in table 8, Appendix C).

Table 4: There are 11,000 Australia based OSL with high likelihood to have the skills for entrepreneurship and are currently working in priority export industries for Sri Lanka Industry Population

Professional, Scientific and Technical Services 2,686 Financial and Insurance Services 1,981 Computer System Design and Related Services 1,870 Tertiary Education 1,487 Education (excl. Tertiary) 1,251 Priority Manufacturing 1,118 Accommodation and Food Service (Potential for Tourism) 748 Total 11,141 Source: Australian Census 2016 e. Conclusion The Australian and US data has showed that the OSL population in both countries is not just highly educated but are also concentrated in managerial, professional, and technical professions often within industries identified as target industries by the BOI. This is encouraging evidence that OSL could play a role in export diversification. Despite the relative concentration in such fields it should be noted that these numbers do not equal the number of potential entrepreneurs. For example, of the 11,000 OSL in Australia employed in professions more likely to lead to entrepreneurship in target sectors, only a proportion will have the right knowhow

and appetite to engage in business activities in Sri Lanka. In the next section we will explore what drives the appetite to do business with Sri Lanka and the barriers to doing so.

4. Case Studies of OSL engagement with Sri Lanka

In this section I outline some prototypical OSL engagement through both the return and overseas channels.

a. Return option

i. Sanjiva Weerawaran, WSO2 Sanjiva left Sri Lanka in 1985 to complete his undergraduate and PhD studies in computer science in the United States, joining IBM research in 1997 after he was finished. While a student he had already started bringing IT to Sri Lanka, having been involved in setting up the first email network to deliver news from Sri Lanka to the US. In 2001 however, with small children growing fast, Sanjiva decided he wanted to move back to Sri Lanka permanently in time for his eldest to start school. He had always wanted to return to Sri Lanka - “it was his home” - and he wanted his children to grow up there. IBM agreed that he could work remotely, and so he was able to make the move with a good job already secured.

Sanjiva soon was involved in the wider Sri Lankan tech scene, starting the Lanka Software Foundation (LSF) to develop open source software, and taking a lectureship position at the university of Moratuwa. In 2004, when the tsunami hit Sri Lanka, he developed (through LSF)

2 the Sahana disaster management software1F to assist the Sri Lankan government coordinate its emergency response.

In 2005, having working remotely for IBM for four years, he left to found WSO2, a software firm that would create and market middleware products centered on the Apache open source software platform he had been working on through LSF. WSO2 was unusual amongst Sri Lankan software companies in that it was creating and marketing a complete product, rather than providing software engineering services to other software companies. Since its founding WSO2

2 The Sahana management system has been deployed in multiple disaster zones since, and is now managed by the Sahana Software Foundation (https://sahanafoundation.org/) 21

has since attracted $45m in venture capital, and now employs 500 people, 460 of whom are based in Sri Lanka (mostly in Colombo, but also Jaffna), with overseas sales staff in target markets (US, Europe, Brazil, Australia). It competes directly with large international IT companies like IBM and serves international clients such as Fidelity Investments.

Sanjiva says he didn’t face too many regulatory or bureaucratic barriers to setting up WSO2. He initially registered the company in Sri Lanka through the Board of Investment (he was able to since he was bringing in foreign exchange) however it became necessary to re-register in the US to reassure investors that they would be able to easily recover their investment. Since the company is registered overseas, tax and trade policy uncertainty (which he believes is a big issue for Sri Lanka), does not affect WSO2. Additionally, as it is a software company, he says it doesn’t rely heavily on Sri Lankan infrastructure or government approvals.

One of the key advantages he feels he had as a returning OSL, was that he better understood the problems of customers in WSO2’s advanced economy markets. Living in Sri Lanka, he says, you can’t feel the issues that international clients face, and the quality demands from local buyers are lower than big firms in the US. Initially venture capital funds asked that he move back to the US and run the company there, while outsourcing the software development to Sri Lanka, but Sanjiva wanted to prove he could build the whole product in Sri Lanka. Sanjiva says he instills a mindset in WSO2 that “we can compete with anybody” and he is passionate about building a world-class software community in Sri Lanka. ii. Pharmaceuticals and Biotech – From Massachusetts, USA to Sri Lanka In 2012 TR Vedananda and a group of other OSL with expertise in pharmaceuticals, biotech, and finance grouped together with the idea of building a pharmaceutical manufacturing plant in Sri Lanka. They had periodically thought “we need to do something” for Sri Lanka, and now they had a plan and the expertise they needed to set up a company. They travelled to Sri Lanka to scope the potential of the idea, which was both encouraging and uncovered an obstacle. If they were going to start a business, they realized they needed to be located in Sri Lanka, but apart from TR Vedananda, who had just retired from a large pharmaceutical company after 25 years and whose children were older and independent, none of the others were able to make the

22 move. They had careers and young families to consider and moving to Sri Lanka was not personally viable.

This story highlights both the opportunity and challenge of starting a pharmaceutical industry in Sri Lanka. The opportunity is large, and can be seen in the story of the founding of the pharmaceutical outsourcing industry in India in the mid-2000s (for more details on pharmaceutical outsourcing see appendix A). Large pharmaceutical companies in North America and Europe were looking to outsource the production of drug compounds which was time consuming and expensive to do in their own laboratories. The Indian outsourced pharmaceutical manufacturing industry began by producing lower complexity compounds used in drug discovery. They slowly built up a reputation for quality until firms were manufacturing at every stage of the process including discovery, clinical trials, and GMP laboratories.

Sri Lanka has two competitive advantages to potentially following the path of India and entering the pharmaceutical outsourced manufacturing industry. First, Sri Lanka boasts a well- educated workforce, with large numbers of biology and chemistry graduates (partly due to the large number of students who aspire to be doctors and tailor their A levels in such a direction) (interview with university professor – interview A). Second, there is a large OSL community in Massachusetts (approximately 1,400 – see earlier ACS analysis) with many working in the pharmaceutical and biotech industry (interview A). The challenge is of course catalyzing a team of OSL to return to Sri Lanka and complete what TR Vedananda had planned. b. Overseas option i. Savan Raknish, PR, design and marketing agency Savan Raknish runs a PR, Design and Marketing agency with his wife based in Singapore but with clients across south east Asia, Japan and Oman. The agency has an office in Colombo with approximately 10 employees that do some of the back-office design work that can be done remotely. Savan describes such an operation as a “design Kade”, a reference to the small local stores called Kades which sell groceries and other items. He says there “are dozens of people [OSL] doing this”, outsourcing small parts of their digital business e.g., running their Facebook adverts out of Sri Lanka.

Savan says he opened the office in Sri Lanka because of the competitive wages for design and web development professionals in Colombo, which are lower than big IT hub cities in India. He also says that Sri Lankans have a good understanding of the “western aesthetic” that lots of his clients are looking for, and that makes them well suited to design work. That Savan is Sri Lankan and speaks the language and understands the culture makes things easier too. He can easily assess a CV and can find someone through his personal networks to act as a supervisor for the office. This cultural understanding and personal connection means its “harder for them to rob me”, in comparison to setting up in a country one has no connection to.

Another aspect to opening the office is altruistic however, he says he “feels it’s his duty to do so”, and he says that many other similar entrepreneurs feel similarly, that it’s their duty. They think “let’s get someone from Vietnam, oh no let’s not, let’s get someone from Sri Lanka”.

There are some disadvantages to being located in Sri Lanka however. He says the quality of training in software development is good but not the best, and that they are best suited to mid- range development tasks. And while spoken English is good, written English is not, and this is what is most useful for the kind of work his office does. Aside from the skills of the workforce he also cites the restrictive labor laws, that make it very difficult to fire permanent employees, which disincentivizes him from hiring employees permanently to avoid being stuck with a poor- performing employee. Additionally, bringing in foreign exchange to pay his employees is often subject to bureaucratic delays (despite the regular modest sums involved) due to central bank checks. Savan also mentioned that other “design kades” suffer from not being able to use PayPal in Sri Lanka to send money. A final issue is passport weakness, which prevents employees travelling to work with at client locations. Savan says he can “deploy Malaysian staff to Singapore, Thailand even Japan with ease – but I can’t do this with Sri Lankan staff, even Cambodian staff have better mobility.”

These challenges have led his business partners to suggest relocating the office to Vietnam, Malaysia or the Philippines (where written English is very good). Savan says the altruistic element is probably the deciding factor as to why he keeps his office in Sri Lanka.

ii. Krishan Canekeratne, Virtusa Krishan’s founding of Virtusa was recently chronicled in “Cultural Transformations : Lessons of Leadership and Corporate Reinvention from the C-Suite Elite” (Mattone and Vaidya, 2016). Krishan left Sri Lanka to study for his bachelors in computer science in the US, graduating in 1988. He worked in the IT industry for several years before co-founding Virtusa with his wife Tushara Canekeratne in 1996. They began with “with six engineers working from the basement of his house in the suburbs of Boston, and another 10 working from his parents’ home in Colombo”(Mattone and Vaidya, 2016).

Having survived the dot-com crash in 2001, Virtusa grew exponentially with an average annual growth rate of 23 percent. Virtusa continued to operate from the US, moving out of Boston to Westborough, Massachusetts. In 2015, Virtusa had revenues of $478m, over 10,000 employees, was the largest tech employer in Sri Lanka. It has also expanded to the UK, the Netherlands, Germany, Sweden, Hungary, Malaysia, Singapore and is one of India’s top 10 tech employers (Mattone and Vaidya, 2016).

5. Lessons for Policy from Current OSL Engagement

The first question to ask is whether the case studies above and other interviews conducted are consistent with the analytical framework set out at the beginning of this policy analysis. WSO2 provides an excellent example of an entrepreneur OSL using the knowhow they had gained working in the United States coming back and starting a firm with that knowhow. With Sanjiva’s knowhow WSO2 could immediately compete with large US firms, something that would normally take years of industry development to achieve. Savan Raknish’s design “kade” provided an excellent example of the game theoretic model outlined in the overseas option. Part of the appeal of opening an office in Sri Lanka was that it “was harder for them to rob me”, exactly the assumed advantage that OSL have when opening an office in Sri Lanka but being based in a foreign market.

Having found the experiences of OSL entrepreneurs consistent with the analytical framework found in the literature we move to considering the two issues most relevant for designing a government policy to encourage more OSL engagement, that of motivation for engagement

and barriers to engagement. Understanding why OSL decide to engage is critical for designing a policy that leverages latent motivations. Understanding the barriers that hamper them from realizing aspired engagement is critical to design policy that minimizes those constraints. It is through understanding these issues the GoSL can help catalyze, for example, some of the 11,000 potential entrepreneurs in Australia into actual entrepreneurs generating exports for Sri Lanka.

a. Motivations and Catalysts

i. Post-war reconstruction As we saw earlier, a large proportion of the OSL population is of Tamil ethnicity, many of whom left the country due to the civil war. With the civil war over, there remains much reconstruction to be done and livelihoods to be developed for war-affected communities. A member of the Tamil OSL community, Naga Narendran, returned to Sri Lanka to start an NGO to help communities develop livelihoods using the skills he gained living and working in Australia for 30 years. He was motivated by wanting to help communities affected by the war in a region he grew up in (interview I – a fuller account of his story is detailed in Appendix E). There are many post-conflict issues to be addressed and for OSL who have connections to the northern and eastern provinces, and this can be a strong motivator for engaging with Sri Lanka.

While this kind of engagement can clearly be of immense value to the war-affected communities, it is unlikely to result in the export diversification that is the goal of this policy analysis. Naga Narendran was able to improve the livelihoods of farmers immensely by organizing them around a common crop, but it wasn’t focused on the export market. The war has left the northern and southern provinces with limited infrastructure, making them some of the more challenging parts of Sri Lanka to set up an internationally competitive export business of the kind identified by the BOI.

ii. A sense of duty It was not just OSL from war-affected regions that felt a desire to “give back” however. Nearly all of the OSL interviewed (full list in Appendix A) expressed the sentiment, that if they could do something they would. An Australian business executive described “knowing how far $50 can

go in Sri Lanka, we had to do something now” (interview B). An American business executive talked of their “belief that something had to get done, given what [Sri Lanka] has been through” (interview C). An American university professor described having “come through the free education system … I always had the feeling that I’ll help, some way I can” (interview A). Savan Raknish and his design Kade and TR Vedananda and his pharmaceutical manufacturing aspirations both had a sense of duty to Sri Lanka (interview G, H).

There is however a selection bias in those interviewed, since the selection process identified OSL who had engaged or considered engaging, and two interviewees did question whether OSL felt as strong a sense duty to Sri Lanka as for example Indian diaspora felt a duty to help India. Nonetheless, it is clear that it is not just financial rewards that have driven people to engage with Sri Lanka, and the government should leverage this fact.

iii. Family and cultural ties Many OSL have family that still reside in Sri Lanka and this can be a significant draw to the country, especially when parents become elderly and need support. Having family present can also result in OSL returning to Sri Lanka regularly for vacations (and potentially maintaining networks with school friends etc.). This was cited as being an important factor when deciding to engage economically with Sri Lanka, as a useful familiarity with the country and local networks are maintained (interviewee C).

There are also often strong cultural connections that draw OSL back. Mohamed Munas, a Sri Lankan diaspora researcher (interview F) noted that retirees were particularly likely to return to Sri Lanka for the cultural and familial connections. No doubt they are also drawn by the high standard of living they can achieve with pensions and savings earned in high-income countries. Several OSL spoke of considering spending more time in Sri Lanka once their children had left school or they were nearing retirement (interview B, I, E). Other OSL who had returned said they were motivated by a desire that their children grew up in Sri Lanka (interview L, M) or simply that Sri Lanka was their home, and that they wished to live there if they could (Interview N, K). Nayana Samaranayake, founder of WorkInSriLanka, an organization set-up to help returning OSL settle in Sri Lanka, stated that “99% of people moving back were due to family,

you had to take over the family business, you had family in Sri Lanka, or you wanted your kids to grow up in Sri Lanka” (interview N). iv. Engagement catalysts One thing that was common across most of the interviews was the presence of a catalyzing moment that led them to engage with Sri Lanka. One spoke of a request made through a government minister for their assistance (interviewee A), another spoke of returning to Sri Lanka for a business conference which prompted him to reconnect with a school friend now in business (interviewee C). Another had a school friend reach out directly from Sri Lanka asking for assistance with a medical facility he was hoping to build (interviewee B). The government should keep this in mind when designing policy to engage OSL, that they can play a role in catalyst events that spark economic engagement.

Push factors from their overseas location can also prompt OSL to return and engage with Sri Lanka. One OSL mentioned that he had returned to Sri Lanka from the US partly due to the dot- com crash and the events of 9/11 increasing the difficulty of obtaining visas (interview K). Saxenian notes that there was a widespread return of Silicon Valley based diaspora at this time. As noted previously, India saw an influx of returning citizens, but China too saw many return in a phenomenon dubbed B2C (back-to-China), the Chinese news agency at one point in 2002 reported that 130,000 diaspora had returned (Saxenian, 2006).

b. Barriers

i. Motivational barriers For many Tamil OSL who left during the civil war there maintains significant mistrust of the government, and potentially antipathy towards the country. The Sri Lankan civil war was long, claimed many lives, and led many to flee, the memory of which is still present in the minds of large parts of the Tamil OSL community (Van Hear and Cohen, 2017). There are also some OSL who are still working towards the creation of a separate Sri Lankan Tamil state (Van Hear and Cohen, 2017) and thus would be unlikely to assist in the development of Sri Lanka as a country (at least outside the Northern and Eastern regions). For many OSL, the change of government in

2015 marked a step change in attitude towards the government and Sri Lanka, but many Tamil OSL still report safety and security concerns regarding returning to Sri Lanka (interview F).

One of the difficult structural issues with OSL engagement is the lack of employment opportunities with a remuneration level near that of salaries in high-income OSL locations e.g., USA, UK, Australia. This is not of such a large concern for the overseas option, but it does present some difficulties for the return option. Entrepreneurs may wish to spend some time while still in employment to make early explorations of a business idea and reacclimatize to doing business in Sri Lanka before starting their business (interview L). Such a route becomes difficult if there are few highly paid employment opportunities in Sri Lanka, and potentially forces entrepreneurs to have a larger quantity of start-up capital to fund their living expenses in the early phases of the business.

Sanjiva Weerawarana was fortunate to be able to work for IBM remotely and thus continue receiving an income before he started WSO2 (interview M), but not many OSL have such an option. Mano Sekaram reported it being one of the biggest difficulties he had to overcome when he returned to Sri Lanka in the early 2000s. He eventually found a role, which enabled him to settle himself and his family in Sri Lanka before starting 99x, now a highly successful software engineering exporter (interview L). Nayana Samaranayake (of WorkInSriLanka), agreed that returning OSL he worked with often find it a barrier, especially for anyone with more than 3-4 years’ experience. While firms are often interested once introduced to candidates, they don’t often advertise for management roles suitable for experienced returning OSL, which can make finding employment difficult (interview N).

Another difficulty for the return entrepreneurship option are the difficulties of moving a family back to Sri Lanka. The most common age for entrepreneurship is considered to in one’s 40s and early 50s (Parker and Parker, 2009). This is also an age when OSL are likely to have young children who have strong ties to the overseas country they have largely grown up in. Uprooting one’s family from the UK or Australia to come to Sri Lanka can be difficult. Such an issue was the core reason why TR Vedananda and his OSL associates chose not in the end to start their pharmaceutical manufacturing operation in Sri Lanka (interview G). It is noteworthy that nearly

all the OSL return entrepreneurs in the private sector arrived without children, or with young children (interviews K, L, M, N).

ii. Technical barriers One of most frequently cited issues is immigration. A study in 2013 found only a few thousand OSL were dual citizens (Reeves, 2013) and the process for claiming it was identified as a “long tedious process” (Young Political Leaders’ Forum and International Alert, 2016). This can cause significant difficulties when trying to return. A researcher from the Centre for Poverty Analysis (CEPA) spoke of the challenge of getting visas for an OSL conference hosted in April 2017, even though the Ministry of Foreign Affairs (MoFA) was supporting the conference and handled the visa process (interviewee F). It no doubt would have been more difficult without MoFA support. Part of the issue is the government security screening process which barres some OSL from returning. This increases distrust among OSL (see above section) and creates a sense of insecurity that may prevent OSL from engaging with Sri Lanka.

Even for those who pass security screening, they can often only get a 30-day tourist visa. And despite dual citizenship being at least a possibility (where their first country of citizenship allows it) an Australian OSL expressed concern about taking dual citizenship, believing it might affect their ability to gain entry to other countries that require visitors to declare dual citizenship (interviewee E). The March 2018 announcement from the government regarding the creation of a special lifetime visa for OSL could significantly ease these problems, although the details are yet to be announced, including whether the degree of security checks will change, and thus whether it will solve the issues highlighted.

Another cited issue is foreign exchange. Controls on foreign exchange are reflected in the Heritage Foundation’s Economic Freedom Index which ranks investment freedom in Sri Lanka at 35, substantially below the average for the Asia-pacific region of 44.2 (The Heritage Foundation, 2017). An Australian business man discussed his concern about setting up a business or investing in Sri Lanka due to foreign exchange controls (interview B). As Savan Raknish described in the case, even bringing money into the country can be difficult with many bureaucratic delays for relatively small amounts of money (interview H). Interview B agreed but

expressed greater concern about getting money out again. If you do not think you will be able to take profits out of the country, that will bring significant caution to investment decisions, even if you are planning on returning to Sri Lanka permanently. One OSL entrepreneur wanted to bring in an American co-founder in the early 2000s to act as sales agent in the US and give him a small stake of equity but was refused due to exchange controls (interview K).

Dealing with administrative requirements for setting up a business, getting permits etc. is a problem that many OSL have faced when trying to start organizations in Sri Lanka (interview C, E, F, N), and those that have worked with returning OSL believe it is a particular issue for them since they have less familiarity with the latest rules and regulations than other Sri Lankans (interview C, N).

Several other issues were mentioned that are general to doing business in Sri Lanka, rather than particular concerns for OSL:

• Human capital: lack of good IT and software professionals (interview H, K, L, M, N) and insufficient English skills (interview H) • Restrictive labor laws that make it difficult to fire permanent staff (interview H) • Wastewater infrastructure for the pharmaceutical industry (interview G) • Policy uncertainty, especially tax policy (interview L, M)

All of these policy concerns were also raised in the growth diagnostic, with the latter two deemed binding constraints. Since these are general business environment issues that are being considered in other forums they will not be addressed further in this report.

6. Return Option vs Overseas Option

I have distinguished the return and overseas options as the two broad channels of OSL engagement that could increase export diversification through private sector entrepreneurship. While many of the policies to encourage the two options would be similar, it is nonetheless valuable for the government to understand which is likely to have the largest effect on diversifying exports, such that it can prioritize policy and measure intermediate progress on that option specifically. I assess the two options along three criteria: technical correctness

(which option, if it occurred would have the largest positive effect on export diversification), administrative feasibility (which option is more feasible given the GoSL’s policy options and capability) and political supportability. I find:

• Technical correctness: return option preferred • Administrative feasibility: equally feasible • Political supportability: equally supportable

Thus, the return option is recommended. Rationale behind the assessment is below.

a. Technical correctness When considering which option, if successfully catalyzed, is most likely to result in export diversification there are two forms of evidence to consider:

1. The form of successful diaspora engagement in export diversification in other countries 2. The experience up until now of OSL’s engagement in the Sri Lankan export sector.

The evidence must then be applied to current context of Sri Lanka.

i. Evidence from other countries Evidence from other countries on the effect of the overseas vs the return option is mixed. The Indian IT industry was mostly aided by the overseas option during the 1990s as they mentored firms and promoted the quality of Indian IT in the US (effectively acting as sales agents). Either side of that period however, the return option looks more impactful. Returning diaspora were pivotal in setting up the first MNC off-shore centers in the 1980s (e.g., HP-India) and Wipro entered the IT industry when a returning diaspora took over the family cooking oil business. In the early 2000s the confluence of the dot-com crash and a burgeoning IT sector drove Indian diaspora back. This cemented much of the growth in the early 2000s, founding at least 200 companies by 2004, and filling senior management role in the MNCs already present (e.g., IBM).

The return option looks to have played a greater role than the overseas option in the growth of the Taiwanese high-tech industry. While much of the industry was seeded through government action and MNCs (e.g., the semiconductor industry), the scanner industry is a clear example of seeding by returning diaspora. Additionally the flood of returnees in the early 1990s drove a 32

proliferation of new startups and filled the management of top Taiwanese firms where they brought the “their experience of managing world-class companies like HP, Intel, and Bell Labs”(Saxenian, 2006). The overseas option was important, with the example of Acer founders relying on former class mates in the US to supply them with knowledge and components. It doesn’t seem to have been pivotal in seeding an industry (Acer was initially responding to domestic demand from an industry that already existed) but instead facilitated Acer’s growth throughout the 80s until it gained the expertise to compete internationally.

The experience of India and Taiwan suggests that the return option can play two roles: industry seeders, and industry consolidators, whereas the overseas option can be critical for developing a nascent industry by providing it with the overseas contacts and market knowhow. ii. Evidence from OSL engagement in Sri Lanka Sri Lanka’s experience of diaspora led growth is of course limited, for if there had been an OSL- led export boom this report would be unnecessary. The IT and BPO export industry is a small story of success however, growing exports from $213m to $850m from 2007-2015 (PWC, 2016). It has provided rich examples for this report and is an obvious example of an export industry OSL could help grow (given the depth of IT knowhow in the OSL community shown in previous sections). OSL engagement in the IT industry has been driven partly by the return option (e.g., WSO2, 99x, Calcey Technologies, CodeGen) and partly by the overseas option (e.g., Virtusa, and a proliferation of “design Kade” type businesses). The rest of the industry growth has been driven by MNCs off-shoring to Sri Lanka (e.g., Pearson, London Stock Exchange, Accenture). iii. Applying the evidence to the Sri Lankan context The Sri Lankan IT sector looks evenly balanced as to whether the return or overseas options are likely to have a bigger impact. The industry potentially resembles the Indian IT industry in the 1990s, established but still small, and could potentially use OSL overseas to act as sales agents to market it as an outsourcing and IT off-shoring destination. Alternatively, the return option might be what’s needed to consolidate the industry. An economic shock to the advanced economies combined with favorable policies to encourage OSL to return to Sri Lanka could lead to an influx of highly skilled OSL into the sector like Indian in 2001-2 and Taiwan in the early

90s. This could establish IT as a major part of Sri Lanka’s exports from the 4% of exports it

3 consists of today2F (Central Bank of Sri Lanka, 2018).

Outside of IT and BPO, the return option looks definitively more promising. The biggest overseas option success in IT was Virtusa, but there’s very little evidence that such overseas entrepreneurship is possible in sectors outside the IT outsourcing business. It is an industry which, by its outsourcing nature, is well suited to starting “with six engineers working from a basement in Boston, and another 10 working from Colombo”. IT is unusual in its limited interaction with infrastructure and government services (interview M). This makes starting a company more possible for someone not physically present in Sri Lanka. In contrast, TR Vedananda and his colleagues concluded that to start a pharmaceutical firm in Sri Lanka they need to relocate permanently to Sri Lanka (interview G). Looking at the list of prioritized BOI industries, this looks more like the norm, with IT and BPO the exception.

What Sri Lanka needs however is new companies seeding new industries. Outside of IT Sri Lanka has little export growth (tourism is the exception), and thus has no existent industry for OSL to promote in the overseas option like the Indian diaspora did for IT in the 1990s. If Sri Lanka needs to seed brand new industries, and if that is not possible for most sectors through the overseas option, then the return option looks a much stronger candidate to facilitate export diversification.

The Taiwanese experience is consistent with this conclusion, with the IC industry requiring government sponsored travel to the US and then returning to set up the first foundry, which seeded the semi-conductor industry. The scanner industry was similarly seeded, but this time organically by returning diaspora, and later came to dominate world markets. Overseas option involvement (e.g., Acer) is most often to facilitate growing industries, not to create brand new ones. If Sri Lanka needs new export industries, the return option is the technically correct channel it needs to prioritize.

3 Latest current account data from 2016 34

b. Administrative feasibility There are clearly fewer intrinsic barriers to motivating an OSL to engage in the overseas option than the return option. OSL in advanced economies are the most settled in their new country of residence and face big wage differentials between their country of residence and Sri Lanka (should they hope to gain employment before starting a company). The difficulty in the return option prevented TR Vedananda from starting the pharmaceutical company because most of his business partners had young children and were settled in the US (interview G). Yet, an advantage of the return option is that while it is more challenging to achieve, one needs just one or two success stories to potentially seed a new industry. Analysis of the Australian OSL community showed that there were 11,000 OSL employed in managerial and professional occupations within the BOI target industries themselves. If just 1% of those chose to return and had the knowhow in a new industry, that would be over 100 new potential seed companies. Additionally, there are more policy levers available to the government when considering the return option, given the much closer contact returnees have with government.

There is also not a significant difference in the capability of the GoSL to implement policies designed to encourage the return or overseas option. The hardest policy challenge in both areas is catalyzing latent motivation to engage, a complex administrative challenge, with large cultural and contextual factors that make importing policies from other countries unlikely to be successful. Removing barriers to the return option (as described in previous section) is likely to be administratively simpler, but these barriers mostly don’t exist for the overseas option. Overall the intrinsic difficulty of facilitating the overseas and return options are broadly equal, and the administrative capability similar and thus administrative feasibility is approximately equal between the overseas and return options.

c. Political Supportability OSL engagement is not without political considerations, given the connection between “diaspora” and the conflict. Nonetheless, the government appears to be engaging in policy reform orientated towards OSL (dual citizenship, and the new lifetime visa) without difficulty, suggesting broad political support for such measures or little political interest in administrative

matters of that type. Policies attempting to catalyze engagement with large amounts of publicity may be more difficult (e.g., OSL festivals). The strong performance in local elections of former President Mahinda Rajapaksa’s new People’s Front party (CNN, 2018) is likely to hamper political supportability. The People’s Front draws its support from the Sinhalese Buddhist population and so engaging with the “diaspora” - often associated with the LTTE - may present a political challenge as the government tries to win back Sinhalese voters. It is unlikely that the political supportability would look particularly different if the intention was to engage through the overseas or return option channel however and thus political supportability is equal between the overseas and return options.

7. Preliminary policy recommendations

Should policy makers agree with the above diagnostic, and that the return option should be prioritized, I outline below a number of policy ideas to be considered that address the barriers and motivations outlined in previous sections. These are again assessed against technical correctness, administrative feasibility, and political supportability. I also assess whether the GoSL is already taking initiative to implement. Table 5 provides a summary, with the four most promising interventions highlighted. Further details follow for the highlighted policies, while details of the less promising interventions can be found in Appendix F.

Table 5: Summary of preliminary policy recommendations Policy recommendation Technical Administrative Political GoSL correctness feasibility supportability initiative? Ease OSL dual citizenship High High High Yes acquisition Visa reform High Medium High Yes Relax capital controls High-Low High Unknown Yes One-stop-shop for OSL High Low High No OSL engagement through High Medium High No embassies and consulates OSL celebration events Medium Medium Low Yes (festivals, conferences)

a. Ease OSL dual citizenship acquisition Technical correctness: high | Administrative feasibility: high | Political supportability: high GoSL initiative: yes Difficulty entering and working in Sri Lanka was identified as a key barrier to engagement. This is because many OSL lost their Sri Lankan citizenship when they applied for citizenship in their new country of residence. The Sri Lankan government has already made it possible to apply for

4 dual citizenship for several years now3F something that the IOM recommends, but initial reports were that it was time-consuming and bureaucratic (International Alert, 2016). The Department for Immigration and Emigration that handles the applications has a strong anecdotal reputation for efficiency (interview O) and has recently upgraded its IT systems suggesting it should be administratively feasible to improve the process. Such a move is clearly politically supportable, since the substantive policy change to allow dual citizenship has already been made.

Next steps: The Prime Minister’s Office should raise the issue with Hon. S.B. Navinne, Minister for Internal Affairs, under whose purview the Department of Immigration and Emigration sits. Performance should be tracked by a) time from application to dual citizenship approval and b) total number of successful for dual citizenship applicants who return permanently to Sri Lanka (the new Dep. of Immigration and Emigration IT system enables such data to be tracked).

b. Visa reform Technical correctness: high | Admin. feasibility: medium | Political supportability: high GoSL initiative: yes Some countries where OSL reside do not allow dual citizenship (e.g., Singapore), and others may be wary of it for fear it could inhibit their travel to other countries (interview E). The Sri Lankan government has very recently announced a new visa category that could allow these groups to still return: an OSL lifetime permanent residence visa (Wickremasekara, 2018). As well as allowing OSL to return without taking citizenship, it also has the benefit of providing easy access to Sri Lanka for OSL exploring the idea of the return option before they go through the administrative requirements of applying for dual citizenship. Setting up a new visa category is likely to add more administrative complexity that improving the existing dual citizenship

4 Although there are restrictions to applying noted earlier, these should not be a hurdle for the vast majority of OSL considering return entrepreneurship. 37

process, hence medium administrative feasibility. The GoSL has already announced the policy showing high political supportability.

Next steps: The Prime Minister’s Office should ensure through partnership with the Minister for Internal Affairs, that the new visa caters to the needs of return entrepreneurs e.g., the right to own property. It should also be quick and easy to acquire to facilitate OSL trips to Sri Lanka while they explore the possibility of return. Performance of the policy should be tracked by a) number of visas granted and b) number of visa holders who settle permanently in Sri Lanka

c. OSL one-stop-shop Technical correctness: high | Admin. feasibility: low | Political supportability: high GoSL initiative: no As described earlier, many OSL spoke of navigating the bureaucracy of Sri Lanka a significant barrier to engaging and starting a business. OSL are likely particularly hampered by this compared to locals given their relative unfamiliarity and lack of connections. For industries outside of IT where interactions with government are more important, this is likely to be a larger barrier. Yet these are the sectors where new firms are most needed. The limited number of successful OSL-led firms outside the IT sector is potentially evidence in itself that this is a big problem. Administratively the BOI is the natural agency to handle such a role, since it already has a mandate to run a one-stop-shop for international investors. A division of this for OSL- specific issues could be created.

Creating a one-stop-shop is a challenging task, given the array of actors involved across the government. Thus, administrative feasibility is assessed as low. Creating such a function is likely to be politically supportable, since it would not require any legislation, and is an administrative issue that is likely to generate much political interest.

Next steps: The Prime Minister’s Office should engage with the Hon. Malik Samarawickrema, minister for Development Strategies and International Trade to initiate the creation of such an office. Performance should be assessed on surveys amongst OSL that use the service.

d. OSL engagement through embassies, high-commissions, and consulates Technical correctness: high | Admin. feasibility: medium | Political supportability: high GoSL initiative: no Most OSL who have taken the return option spoke of a catalyst that initiated the move. Many of these were often due to push factors (e.g., US recession) or personal pull factors (e.g., family) that the government can do little to influence (although it should be ready for a potential advanced economy recession). One spoke of the key role of the Australian high-commissioner however, who both encouraged him and connected him with government officials who could facilitate his venture (interview I). Another Australian OSL spoke of the high-commissioner’s active role in the OSL community (interview E). The network of diplomats situated in each of the key OSL countries represents the most proximate contact between the government and OSL, and thus an excellent channel to catalyze OSL engagement.

The Ministry of Foreign Affairs faces a capability shortfall however, with only 183/264 of the

5 staff officer positions filled to cover a total of 67 positions4F and many struggling to command English, let alone other foreign languages (Alphonsus, forthcoming). Staff shortfalls have been overcome by appointing high-commissioners from outside of the foreign service, a positive result of which has been the capability of the Australian high-commissioner. That such an option exists, and the prestige of the positions in the target countries (US, Aus., Canada, UK) that makes it likely strong external candidates will apply, rescues administrative feasibility to a medium ranking. There is little reason to see why the outreach by ambassadors to citizens overseas should enter the political intercourse, and thus political supportability is high.

Next steps: The Prime Minister’s Office should engage with the Minister of Foreign Affairs, the Hon. Thilak Marapana to devise a strategy for engagement with OSL overseas in collaboration with the BOI. More analysis of census data of the like performed in this report can be done to help target engagement towards areas with high numbers of OSL who work in promising industries. Connecting the work of the BOI’s pitches for investment in certain sectors with the MoFA’s OSL networks could be particularly fruitful in attracting return OSL in promising export industries. We also learnt that many OSL return to bring their children up in Sri Lanka or after

5 Compare this to Singapore’s 867 officers to cover 50 missions (Alphonsus, forthcoming) 39

retirement, and embassies should thus focus on OSL in those age groups. If there are staff and resource constraints, the findings of this report suggest there is likely to be motivation within the OSL community to volunteer their time to assist engagement (especially retired OSL).

Performance of the policy will be difficult to assess, as identifying causal influence for something like the decision to return is challenging. Qualitative assessments, through conversations with OSL in the relevant countries will be the best assessment of effectiveness.

Appendix A – Interview list

Code Current Return/ Name Organization Description Location Overseas OSL A USA Overseas Anon. Anon. University professor Business executive in the IT B Australia Overseas Anon. Anon. sector Business executive in C USA Overseas Anon. Anon. financial services Pharmaceuticals D USA Overseas Anon. Anon. professional Entrepreneur in the E Australia Overseas Anon. Anon. import/export business

Mohamed Centre for Poverty Researcher at Centre for F Sri Lanka N/A Munas Analysis Poverty Analysis (CEPA) T.R. Pharmaceuticals G USA Overseas Vedananda Novartis (retired) professional Graphic design company PR, design and marketing H Singapore Overseas Savan Raknish (Singapore/SL) agency owner Institute of Regional Naga Development and Business professional and I Australia Return Narendran Governance NGO founder Sri Lanka (previously J Australia) Return Anon. Anon. Business owner Sri Lanka (previously Mangala Technology business K USA) Return Karunaratne Calcey Technologies founder Sri Lanka (previously Mano Technology business L Australia) Return Sekaram 99x founder Sri Lanka (previously Sanjiva Technology business M USA) Return Weerawarana WSO2 founder Nayana Storylize and Technology business N USA Return Samaranayake WorkInSriLanka founder Current policy maker in O Sri Lanka N/A Anon. Anon. GoSL Former policy maker in P Sri Lanka N/A Anon. Anon. GoSL Former policy maker in Q Sri Lanka N/A Anon. Anon. GoSL

Appendix B – Pharmaceutical outsourcing

Pharmaceutical outsourcing can occur at several different stages of the process from drug discovery to manufacture. The simplest production outsourcing occurs at drug discovery, when after modelling a new compound that has potential to treat a particular disease, enough quantities are required to perform initial testing on animals such as rodents. The level of complexity in manufacturing is closely linked to the purity of the final compound that is required. Initial testing on rodents requires much lower levels of purity, and thus is easier to perform than later stages on primates and eventually human’s clinical trials. Thus, this initial stage was the first to be outsourced, as the degree of expertise and proved track record required was lower. Similar lower complexity manufacturing can take the form of drug intermediates. These are input compounds that require further processing to arrive at a final product, and also have lower purity requirements since further refining and processing is will occur in-house in the purchasing pharmaceutical company.

The most complex stage to be outsourced is the final manufacture of the drug for human consumption. These are known as Good Manufacturing Practices (GMP) plants and produce pharmaceuticals of the highest purity to ensure the safety of consumers and are also the most complex part of the pharmaceutical manufacturing industry.

Source: interviews with pharmaceutical professionals (interview D, G)

Appendix C – OSL Characteristics in USA and Australia

Table 6: Most US based OSL are located in highly developed, urban centers State/Area Population Percentage of total California 10,497 27% New York / Connecticut / New Jersey 7,558 19% Texas 5,394 14% Washington DC / Virginia / Maryland 4,089 10% Massachusetts 1,378 4% Other 10,132 26% Total 39,048 100% Adult populations of OSL by state/area in the US. Source: American Community Survey (ACS) 2016

Figure 6: Most US based OSL work in managerial and professional roles

Number of OSL by occupation in the US

Management, Business, Science, and Arts Education, Training, and Library Computer and Mathematical Healthcare Practitioners and Technical Sales and Related Occupations Office and Administrative Support Transportation and Material Moving Personal Care and Service Life, Physical, and Social Science Financial Business Operations Food Preparation and Serving Other (11 occupations) - 1,000 2,000 3,000 4,000 5,000 6,000

Number of OSL in each occupation in the US. Source: American Community Survey (ACS) 2016

Table 7: Australian OSL are heavily concentrated in Melbourne and Sydney State/Area Population Percentage of total Greater Melbourne 54,032 49% Greater Sydney 26,572 24% Greater Perth 7,284 7% Greater Brisbane 7,086 6% Greater Adelaide 3,517 3% Australian Capital Territory 2,774 3% Other 8,592 8% Total 109,857 100% Adult populations of OSL by city in Australia. Source: Australian Census 2016

Figure 7: Australian OSL are concentrated in professional and managerial roles

Number of OSL by occupation in Australia

Professionals 20,930 Clerical and Administrative Workers 9,708 Labourers 8,248 Managers 7,059 Technicians and Trades Workers 6,319 Community and Personal Service Workers 6,188 Sales Workers 3,654 Machinery Operators and Drivers 3,359

- 5,000 10,000 15,000 20,000 25,000

Number of OSL in each occupation in Australia. Source: Australian Census 2016

Table 8: Australia based OSL in occupations suited for entrepreneurship and spread across a range of priority export industries for Sri Lanka Industry Population

Professional, Scientific and Technical Services (except Computer System Design and Related Services) 2,686 Computer System Design and Related Services 1,870 Tertiary Education 1,487 Financial and Insurance Services 1,415 Education (excl. Tertiary) 1,251 Auxiliary Finance and Insurance Services 566 Food and Beverage Services 504 Food Product Manufacturing 314 Machinery and Equipment Manufacturing 268 Accommodation 244 Transport Equipment Manufacturing 174 Polymer Product and Rubber Product Manufacturing 106 Fabricated Metal Product Manufacturing 77 Primary Metal and Metal Product Manufacturing 58 Textile, Leather, Clothing and Footwear Manufacturing 52 Non-Metallic Mineral Product Manufacturing 46 Furniture and Other Manufacturing 23 Total 11,141 Source: Australian Census 2016

Appendix D – Entrepreneurially likely occupation categories in Australian census data

Table 9: The occupations most likely to engage in entrepreneurship are those already in managerial or professional roles Professions Number of OSL Chief Executives, General Managers and Legislators 499 Specialist Managers 4,398 Hospitality, Retail and Service Managers 1,885 Managers (not further defined) 211 Business, Human Resource and Marketing Professionals 5,857 Design, Engineering, Science and Transport Professionals 3,371 Education Professionals 1,968 ICT Professionals 3,778 Professionals (not further defined) 426 Total 22,393 Source: Australian Census 2016 Appendix E – Additional case study

Naga Narendran, Institute of Regional Development and Governance (IRDG)

Naga grew up in Jaffna, Northern Sri Lanka and moved to Australia in 1984. He had a long and successful career in retail and then as a management consultant specializing in supply chain management. After the new government arrived in 2015 he felt it was the right environment for him to help in the post-war development of the northern and eastern provinces. He spent 18 months out the next two years in Sri Lanka on a volunteer basis, gaining an understanding of the development issues in the region, and working to set up IRDG with a former civil servant as CEO.

IRDG designs and implemented private sector-based development project. These projects help villagers cooperate on agricultural production, enabling them to produce one product at a larger scale (known as One Village One product). IRDG then helps connect the villages to agro- byers to sell their product. For example, in Kilinochchi a village has started large-scale free- range organic chicken farming. Having so many farmers in one area means they can procure 46 common services such veterinary services and cage maintenance at much lower cost (something that IRDG facilitated). The net result was to halve the cost of production from the small-scale individual production before. The village will be connected to grocery chains in Colombo, who were importing chicken and eggs, but prefer to have local suppliers. Naga used his business expertise to do the value-chain and financial analysis required to make the enterprise profitable, and attractive to large agri-business buyers.

Naga says he was motivated by the chance to help people from a region where he came from. He had gathered a lot of skills during his career, and he could have continued to work in Australia, but to take the skills he learnt in business and his language knowledge (he speaks English, Sinhala and Tamil) and apply them in Sri Lanka was far more challenging, and personally satisfying.

When Naga initially had the idea, he approached the high-commissioner in Australia who put him in touch with an advisor in the Prime Minister’s office. This was a very helpful connection and provided advice and encouragement at the initial stages. Ultimately, Naga set up the organization largely independently from the government, but since its success they have begun to support it, with government funding to work with more villages in the 2018 budget.

Having successfully setup IRDG, with local staff hired and based in Killinochchi, IRDG now uses a virtual office system to connect OSL outside Sri Lanka with the IRDG team. The expansion of internet services to rural parts of the region has enabled this remote working model. He was invited to share his experiences in the 2017 OSL roundtable organized by the MoFA and CEPA. His observations were that OSL were motivated to do similar things, but that frustrations such as getting visas created barriers to action “as soon as it is hard work, it becomes less likely”.

Appendix F – De-prioritized preliminary policy recommendations

a. Capital controls Technical correctness: high-low | Admin. feasibility: high | Political supportability: unknown GoSL initiative: yes Capital controls of various forms were described as a barrier to engagement e.g., bringing foreign exchange in to pay workers (interview H) and (in the mid-2000s) offering equity to non- residents (interview K). The government has already made efforts to relax capital controls with a new foreign exchange law that came into effect in November 2017 (Sunday Times Sri Lanka, 2017) that also laid provisions specifically for OSL and may have resolved some the issues raised in this research. Easing capital controls has wider macroeconomic effects that are beyond the scope of this paper, so caution should be taken in changing laws purely for the benefit of OSL (hence the wide range of technical correctness). Administratively implementing the regulations should not be difficult given they can be enacted by a small group within the highly capable central bank (author’s experience). Political support for the latest round of capital controls reform was clearly present, support for further reform is unknown.

Next steps: The Prime Minister’s Office should engage with CBSL determine whether a) the latest capital control reforms are sufficient to address the issues raised in this paper and b) understand the macroeconomic risks of further reform if they have not.

b. OSL celebration events Technical correctness: medium | Admin. feasibility: medium | Political supportability: low GoSL initiative: yes Catalyzing OSL engagement could be achieved through holding festivals and conferences for OSL. Returning for a conference was the catalyst for one OSL engaged in the overseas option (interview C). As previously noted the MoFA has supported the organization of third-party conferences for OSL, but it has not engaged in organizing events itself. The GoSL could host a celebration event, inviting OSL back to Sri Lanka and celebrating the success of OSL. This aim would be to catalyze OSL returning to Sri Lanka by introducing them to networks and tapping into latent altruistic motives, an idea also endorsed by The Roadmap report (Young Political Leaders’ Forum and International Alert, 2016).

While a conference can could inspire OSL, it lacks the long-term engagement that high- commissions and embassies can offer. Administratively, it is understood that the MoFA considered holding a festival for OSL, but this didn’t happen (interview P), which given the resourcing constraints is not surprising. Other ministries could take on the role, although it seems unlikely any have experience hosting such an event. There are also political concerns, with one former policy maker expressing concerns that given the recent electoral success of Rajapaksa, the GoSL may shy away from holding very visible projects to engage with “diaspora” (interview Q).

Next steps: Political supportability should be explored with key members of the coalition before proceeding. Performance would be assessed based on tracking what number of attendees return permanently, and potentially following up to see how many have started industries.

References

Agunias, D.R., Newland, K., International Organization for Migration, Migration Policy Institute, 2012. Developing a road map for engaging diasporas in development: a handbook for policymakers and practitioners in home and host countries. IOM/MPI, Geneva, Switzerland.

Alphonsus, D., forthcoming. Towards a Self-Reforming Foreign Ministry. Lakshman Kadirgamar Institute.

Becker, J., 2006. Funding the" final war": LTTE intimidation and extortion in the Tamil diaspora. Human Rights Watch.

Board of Investment of Sri Lanka, Export Development Board of Sri Lanka, Harvard Center For International Development (CID), 2017. Targeting Sectors for Investment & Export Promotion.

Borjas, G.J., 2014. Immigration economics. Harvard University Press, Cambridge, Massachusetts.

Center for International Development, 2018. Sri Lanka Growth Diagnostic.

Central Bank of Sri Lanka, 2018. Balance of Payments.

CNN, M.S. and S.G., 2018. Sri Lanka government in crisis after comeback of ex-leader [WWW Document]. CNN. URL https://www.cnn.com/2018/02/18/asia/sri-lanka-ex-leader-comeback- intl/index.html (accessed 3.11.18).

Cohen, R., 2008. Global diasporas: an introduction, 2. ed. ed, Global diasporas. Routledge, London.

Constitution of the Democratic Socialist Republic of Sri Lanka, 1978.

Greif, A., 1993. Contract enforceability and economic institutions in early trade: The Maghribi traders’ coalition. Am. Econ. Rev. 525–548.

Hallward-Driemeier, M., Khun-Jush, G., Pritchett, L., 2010. Deals versus rules: policy implementation uncertainty and why firms hate it. National Bureau of Economic Research.

Hausmann, R., Nedelkoska, L., 2017. Welcome Home in a Crisis: Effects of Return Migration on the Non-Migrants’ Wages and Employment.

Head, K., Ries, J., 1998. Immigration and Trade Creation: Econometric Evidence from Canada. Can. J. Econ. 31, 47–62.

Henayaka-Lochbihler, R., Lambusta, M., 2004. The Sri Lankan Diaspora in Italy. Berghof Res. Cent. Confl. Manag.

International Alert, 2016. Engaging Overseas Sri Lankans: Promoting Responsible Investment and Technical Assistance, Conference Report.

Malalgoda, C., Samaraweera, P., Stock, D., 2018. Targeting Sectors for Investment and Export Promotion in Sri Lanka. https://srilanka.growthlab.cid.harvard.edu 185.

Mattone, J., Vaidya, N., 2016. Cultural transformations: lessons of leadership and corporate reinvention.

Mostafa, R., Klepper, S., 2017. Industrial Development Through Tacit Knowledge Seeding: Evidence from the Bangladesh Garment Industry. Manag. Sci. https://doi.org/10.1287/mnsc.2016.2619

OECD, 2015. Connecting with Emigrants. OECD Publishing. https://doi.org/10.1787/9789264239845-en

Pandey, A., Aggarwal, A., Devane, R., Kuznetsov, Y., 2004. India’s Transformation to Knowledge- based Economy: Evolving Role of the. Evalueserve.

Parker, S.C., Parker, S.C., 2009. The economics of entrepreneurship. Cambridge University Press, Cambridge, UK; New York.

Prime Minister’s Office, 2016. Performance Report of the Prime Minister’s Office 2016.

PWC, 2016. SLASSCOM Strategy Document 2016. SLASSCOM.

Reeves, P. (Ed.), 2013. The encyclopedia of Sri Lankan diaspora. Editions Didier Millet in association with Institute of South Asian Studies, National University of Singapore, Singapore.

Rhee, Y.W., Bélot, T., 1990. Export catalysts in low-income countries: a review of eleven success stories, World Bank discussion papers. World Bank, Washington, D.C.

Roundtable on “Overseas Sri Lankans as Partners of Development” [WWW Document], 2017. . comdu.it. URL http://comdu.it/blog/roundtable-on-overseas-sri-lankans-as-partners-of- development/ (accessed 1.28.18).

Ruggles, S., Genadek, K., Goeken, R., Grover, J., Sobek, M., 2017. Integrated Public Use Microdata Series: Version 7.0. https://doi.org/10.18128/D010.V7.0

Samet, K., 2013. Circular Migration between the North and the South: Effects on the Source Southern Economies. Procedia - Soc. Behav. Sci., 3rd World Conference on Learning, Teaching and Educational Leadership 93, 225–242. https://doi.org/10.1016/j.sbspro.2013.09.182

Saxenian, A., 2006. The new argonauts: regional advantage in a global economy. Harvard University Press, Cambridge, Mass.

Sri Lankan Bureau of Foreign Employment, 2016. Annual Statistical Report of Foreign Employment - 2016.

Sunday Times Sri Lanka, 2017. New Sri Lankan foreign exchange laws effective today [WWW Document]. URL https://goo.gl/ZwX1jf (accessed 3.9.18).

The Cabinet [WWW Document], 2018. URL http://cabinetoffice.gov.lk/cab/index.php?option=com_content&view=article&id=12&Itemid=18 &lang=en (accessed 3.10.18).

The Heritage Foundation, 2017. 2017 Index of Economic Freedom.

United Nations, Department of Economic and Social Affairs, Population Division, 2015. Trends in International Migrant Stock: The 2015 Revision, (United Nations database, POP/DB/MIG/Stock/Rev.2015).

US Department of State, 2017. Sri Lanka Investment Climate Statement.

Van Hear, N., Cohen, R., 2017. Diasporas and conflict: distance, contiguity and spheres of engagement. Oxf. Dev. Stud. 45, 171–184. https://doi.org/10.1080/13600818.2016.1160043

Wickremasekara, D., 2018. Sweeping visa policy changes to attract professionals and investors. Sunday Times Sri Lanka.

World Bank Group, 2017. Doing Business 2018. World Bank Group.

Young Political Leaders’ Forum, International Alert, 2016. Roadmap for Engagement with Overseas Sri Lankans.